To provide the greatest listening pleasure, an acoustic loudspeaker system must meet several basic requirements. First, it must be capable of reproducing very low frequencies, such as bass notes below 50 Hz, which are felt, not heard. Second, it must be capable of reproducing overtones of high musical notes. Third, it should have a relatively flat frequency and phase response over the full range of human audible frequencies, from about 40 Hz to about 20,000 Hz in order to reproduce sound with fidelity to the source. Fourth, also to be faithful to the source, the system should recreate whatever spatial illusions are contained in the source material. For example, most music sources are encoded for stereo reproduction using two channels. Two, spatially separated and phase-synchronous infinitesimal point sources of acoustic energy theoretically provide the best stereo imaging, for they are able to create the illusion of sound originating from any point along a line extending through both point sources. Therefore, a loudspeaker system should imitate as closely as possible two infinitesimally small point sources of acoustic energy. Fifth, to accommodate wide dynamic ranges, a loudspeaker system must be able to handle signals with power sufficient to reproduce low frequencies at loud volumes without distortion to the sound or damage to the speaker.
Conventional belief is that a single acoustic driver cannot deliver a frequency range and power handling capability required for high fidelity sound reproduction and demanded by audiophiles. Therefore, to meet these demands, most loudspeaker systems rely on two or more acoustic transducers or drivers per channel. Each driver of a channel is responsible for reproducing sounds in only in preselected portions of the audible range. As more fully explained below, characteristics which optimize an acoustic driver or transducer for high frequency sound are often opposite of those which optimize a driver for low frequency response. By utilizing multiple drivers per channel, each driver may be optimized to operate within a selected portion of the acoustic range. An electrical circuit, known as a cross-over network, splits portions of the energy of the input signal between the drivers, depending on the frequency of the energy in the signal.
Despite their widespread acceptance, multi-driver speakers have several drawbacks. First, cross-over networks distort the electrical sound signal, thus introducing distortion into the sound reproduced by the loudspeaker system. For example, cross-over networks naturally cause phase distortion in incoming signals: higher frequencies will be phase shifted with respect to the lower frequencies. Phase shifting results in a loss of clarity, causing the music to sound "muddy." Cross-over networks therefore sometimes employ complex circuits to correct phase distortion. These complex cross-over networks then often introduce other types of distortion and often possess non-linear responses. Second, multi-driver speaker systems tend to be larger and have more components, thus making them more expensive, bulkier and less mobile. Third, a multi-driver speaker does not satisfactorily represent a point source of acoustic radiation for a single channel, as a channel is obviously radiating from multiple points. Thus, they cannot achieve the best stereo imaging.
Nevertheless, they are still preferred over single driver loudspeaker systems. The problems of using a single driver to reproduce at equal levels high notes with clarity and low notes with physical impact are difficult to overcome. A conventional acoustic transducer has a relatively stiff or rigid diaphragm which reciprocates along a linear axis. For reproducing low frequencies, the diaphragm has preferably a concave, cone shape. For high frequencies, it may be flat or convex. To vibrate the diaphragm, an electrical signal representing the sound wave to be reproduced flows through a coil mechanically connect to the diaphragm. The coil is situated within a fixed magnetic field, causing the coil to reciprocate with changes in the current. The coil is formed from one or more lengths of wire wrapped around a support structure. Typically, the edges of the diaphragm are attached to a basket shaped frame using a compliant, slightly resilient, material. The coil is centered within a gap referred to as a "flux gap," formed between cylindrically shaped pole and a donut-shaped magnet assembly. The prevalent structure for centering the coil within the flux gap is a corrugated cloth impregnated with resin, referred to as a "rear suspension," that extends from coil to the frame.
To provide the most accurate sound reproduction, the movement of the coil in response to the electrical signal and the coupling of the movement of the diaphragm to the air in response to the movement of the coil must be linear. Unfortunately, the responses of these elements to the sound signal are rarely totally linear, especially over the entire audible range. The diaphragm couples the mechanical energy of the moving coil to the air, thereby causing the air to vibrate and setting up acoustic waves. At lower frequencies, the diaphragm can be thought of as behaving like a simple mechanical piston pushing volumes of air. At low frequencies, a lot of power is required to push large volumes of air, particularly at loud volumes. Therefore, to sound low notes with great volume a speaker must be capable of handling a lot of power, particularly the mechanical stresses from the strong electromagnetic forces and resulting heat.
For good low frequency response, a driver is needed which is mechanically strong and powerful in order to move larger amounts of air. Thus, a stiffer diaphragm with a large surface area is preferred. However, a large, stiff diaphragm means more structure, and thus more mass. More mass means less efficiency, and thus more power to reproduce the same loudness. More power means that a more massive coil is required to handle the mechanical and thermal stresses resulting from the power. However, more mass in the moving parts inhibits the driver's ability to reciprocate at higher frequencies. Also, it is more difficult to control coupling of the movement of the coil to the air through a large diaphragm and its natural resonances. A smaller diaphragm could be used to sound bass notes, but a longer throw or stroke of the coil would be required to move the same amount of air. However, a longer stroke necessitates either a magnetic field of greater magnitude or a longer coil in order to provide a sufficiently high electromotive force (EMF). Furthermore, a greater coil length means greater induction. Thus, the length of the coil is limited. A long stroke also requires the coil to move at a higher velocity. Higher velocities will create a higher back EMF, which resists travel of the coil and ultimately limits the ability of the driver to reproduce low frequencies.
At higher frequencies, the diaphragm behaves more like a radiating transmission line. The rapid vibrations of the coil cause not only linear movement of the diaphragm, but also mechanical vibrations in the diaphragm which radiate from the points where the coil is attached, outwardly to the edge of the diaphragm. Depending on the material, size of the diaphragm and how it is attached to the suspension, these vibrations may resonate at certain audible frequencies, thus adversely affecting the linearity of the coupling of the mechanical movement of the coil to the air. Although there may be mechanical deformation of the diaphragm at all frequencies, at high frequencies the effect of resonant vibrations will have a substantial impact on the sound, with certain frequencies being noticeably enhanced and others degraded. Reproducing a high frequency sound also requires the coil to be quickly accelerated. Thus, a near zero mass coil and diaphragm is theoretically ideal. Furthermore, a smaller diameter diaphragm is preferred. A larger diameter diaphragm tends to be more directional, exacerbating the directional nature of high frequencies.
Finally, whether a small or large diaphragm is used, the suspension system must be very compliant to accommodate the range of movement of the coil, yet have enough spring force to keep the diaphragm centered in a neutral position. Compliance is required when sounding low notes in order to avoid interference and damage. A large spring force works against movement of the diaphragm and will tend to bend it. However, a compliant suspension tends to resonate and will not dampen undesirable resonances in the mechanical structure of the diaphragm at higher frequencies, resulting in the suspension vibrating out of phase with the diaphragm and a loss of energy.
Attempts have been made to accommodate the demands of high and low frequencies in a single, broad band acoustic driver, particularly in the area of reducing the mass of the moving parts of the driver. For example, as shown in U.S. Pat. Nos. 4,115,667 and 4,188,711 of Babb, the conventional rear suspension for the coil is replaced with a low friction bearing made of TEFLON.RTM.. The bearing is formed at the bottom of the coil, opposite of where it connects to the diaphragm, and encircles and rides on the post. The coil remains centered within the gap without the extra mass of the rear suspension and its spring forces interfering with movement of the coil. The coil therefore can move more freely and accelerate faster, which aids in moving the coil long distances when using a longer throw coil to sound bass notes. Lightweight, stiff metal alloys have been used to form diaphragms. Coil forms (structures for supporting windings of coils) have been made from high strength, thermally resistant materials such as KAPTON.RTM.. To provide a low mass, compliant suspension for the diaphragm, a stamped synthetic foam having a very low density with good dampening and resonance characteristics is used.
Nevertheless, although not recognized in the art, there still exist problems. First, a coil undergoes great mechanical stress from the EMF generated by the magnet and the current running through the coil, as well as great thermal stress from the substantial heat generated when large currents flow through the coil during reproduction of loud notes. Despite the use of lightweight, stiff materials, a low mass coil capable of sounding both high and low frequencies will naturally tend to be weaker and thus more easily deformed by the mechanical and thermal stresses suffered during reproduction of high power sounds. A deformed coil cannot sound notes as accurately and will tend to rub against the walls defining the flux gap, causing noticeable distortion of low notes and extraneous noises at midrange frequencies. Second, a low mass coil also cannot store heat for later dissipation. Thus, during extended periods of loud notes, a low mass coil will tend to get very hot and become damaged. Furthermore, TEFLON.RTM. is not structurally strong and tends to shrink in heat, thus resulting in increased drag of the coil's bearing on the post and deformation under high thermal and mechanical loads. Third, a low density suspension is relatively transparent to sound. Thus, acoustic energy directed rearwardly into the enclosure in which the driver is mounted will leak through the suspension, resulting in sound which is slightly murky due to the delay in the reflected sounds mixing with the sound emanating directly from the driver. Fourth, the suspension, due to large excursions, becomes fatigued where it joins the diaphragm and the frame. Fifth, the thin metal used to form a diaphragm still bends, creating a non-linear response, and eventually becomes fatigued.